|
When Google released Gemini 3 Pro at the end of last year, it was a significant step forward for the company's proprietary large language models. Now, the company is bringing some of the same technology and research that made those models possible to the open source community with the release of its new family of Gemma 4 open-weight models.
Google is offering four different versions of Gemma 4, differentiated by the number of parameters on offer. For edge devices, including smartphones, the company has the 2-billion and 4-billion "Effective" models. For more powerful machines, there's the 26-billion "Mixture of Experts" and 31-billion "Dense" systems. For the unfamiliar, parameters are the settings a large language model can tweak to generate an output. Typically, models with more parameters will deliver better answers than ones with less, but running them also requires more powerful hardware.
With Gemma 4, Google claims it's managed to engineer systems with "an unprecedented level of intelligence-per-parameter." To back up this claim, the company points to the performance of Gemma 4's 31-billion and 26-billion variants, which claimed the third and sixth spots respectively on Arena AI's text leaderboard, beating out models 20 times their size.
All of the models can process video and images, making them ideal for tasks like optical character recognition. The two smaller models are also capable of processing audio inputs and understanding speech. Separately, Google says the Gemma 4 family is capable of generating offline code, meaning you could use them to do vibe coding without an internet connection. Google has also trained the models in more than 140 languages.
|
|
Save 30% with a OnePlus coupon this month, plus save up to 10% on earbuds, phones, and more.
|
RELATED ARTICLES | | |
|
Discover how to save on Ring cameras, doorbells, outdoor cameras, and more.
|
|
What should have been a routine release has revealed some of the features Anthropic has been working on for Claude Code. As reported by Ars Technica, The Verge and others, after the company released Claude Code's 2.1.88 update on Tuesday, users found it contained a file that exposed the app's source code. Before Anthropic took action to plug the leak, the codebase was uploaded to a public GitHub repository, where it was subsequently copied more than 50,000 times. All told, the entire internet (and Anthropic's competitors) got a chance to examine more than 512,000 lines of code and 2,000 TypeScript files.
In the aftermath, some people claim to have found evidence of upcoming features Anthropic is working to develop. Over on X, Alex Finn, the founder of AI startup Creator Buddy, says he found a flag for a feature called Proactive mode that will see Claude Code work even when the user hasn't prompted it to do something. Finn claims he also found evidence of a crypto-based payment system that could potentially allow AI agents to make autonomous payments. In a Reddit post spotted by The Verge, another person found evidence that Anthropic might have been working on a Tamagotchi-like virtual companion that "reacts to your coding" as a kind of April Fools joke.
"A Claude Code release included some internal source c
|
|
